Information Coverage and Truthfulness Evaluation on Web-based Retrieval
0.749S_fact ScoreMixtral-8x22B
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| Mixtral-8x22BAlignment LLM=Llama-3.1-70B2025.01 | 0.749 | — | — | 0.503 | 0.601 | 0.429 | 0.545 | |
| GPT-4Alignment LLM=Llama-3.1-70B2025.01 | 0.748 | — | — | 0.551 | 0.634 | 0.541 | 0.627 | |
| Openchat 3.5 (7B)Alignment LLM=Llama-3.1-70B2025.01 | 0.741 | — | — | 0.52 | 0.611 | 0.444 | 0.555 | |
| Llama-3-70BAlignment LLM=Llama-3.1-70B2025.01 | 0.714 | — | — | 0.556 | 0.625 | 0.486 | 0.578 |